Origins, philosophy, and the case for structured prioritization in personal health.
The longevity space has more information than any individual can act on. Hundreds of interventions, dozens of wearables, thousands of experts — all competing for attention. What is missing is not more data, but structure. The Asymptote Principle is a framework for prioritizing personal health interventions using one observation: every intervention follows a curve of diminishing returns, and the steepest part of the curve is where the biggest gains live. This paper sets out where the idea came from, what it implies, and why we believe that structure applied to existing science is, in itself, the contribution.
In 1975, the demographer Samuel H. Preston published a landmark observation. He plotted the relationship between national GDP per capita and life expectancy and found a characteristic shape. At low income levels, small increases in wealth produced dramatic gains in life expectancy — through better sanitation, nutrition, basic healthcare, and housing. But beyond a certain threshold, additional wealth bought progressively less additional lifespan. The curve flattened.
Preston was describing populations, not people. But the shape of his curve — the principle of diminishing returns — turned out to be almost universal. Education, training intensity, dietary improvements, sleep duration, social connection: every domain of human optimization seems to follow a similar saturating curve. The first steps are large. Each subsequent step is smaller. The asymptote — the theoretical maximum — is approached but never reached.
The leap from Preston's macro observation to personal health is short, but its consequences are large. If every health domain follows a saturating curve, then what to do first is a different question for every person. A sedentary person and a marathon runner sit at very different points on the exercise curve. A chronic insomniac and someone with a steady seven-and-a-half-hour sleep schedule sit at very different points on the sleep curve. The next intervention with the largest expected return depends entirely on where you currently sit.
This is what generic advice misses. “Eat more vegetables” is good advice for most people, but it is not the first thing for someone sleeping five hours a night. “Try this new biomarker test” is interesting, but it is not the first thing for someone who hasn't gone for a walk in a month. The information industry treats all interventions as equal. The curves say they aren't — at least not for any specific person, at any specific time.
Three implications follow.
First, not all interventions are equal — and the inequality depends on who you are. The same advice is worth a great deal to one person and almost nothing to another, even within the same evidence-based intervention. The variable is position on the curve.
Second, the biggest gains are always at the bottom. Going from terrible to mediocre in any domain produces more healthspan benefit than going from good to excellent. This is unintuitive in a culture that celebrates optimization, but it is mathematically straightforward.
Third, there is a natural sequencing. Identify the steepest curve. Climb it until marginal returns diminish. Then move to the next. One habit at a time.
Most of what Asympt knows, the literature already knew. The dose-response relationships for exercise, sleep, nutrition, social connection, stress, screen use, fertility, and emotional skills have been published in peer-reviewed journals for decades — in some cases for more than half a century. The science exists. What did not exist was a way for an individual to use it.
The structure is the product. A framework that converts thousands of pages of population-level studies into a single, personally-relevant answer to the question “what should I do first?” is not novel science. But it is something that did not previously exist as a thing a person could actually use.
This is also why the methodology is deliberately open. Curves, weights, and the evidence each curve is anchored to are visible. The intent is not to obscure what is underneath but to make it usable. Anyone who wants to inspect the equations can; anyone who just wants the answer to the question can have that too.
Asympt is a free assessment platform. It currently spans four pillars of modern human flourishing — longevity, digital wellbeing, fertility, and emotional readiness — chosen because each represents a distinct dimension of how a person navigates a modern life: physical, attentional, reproductive, and emotional. Each pillar applies the same asymptote methodology, identifying the user's position on each curve within that domain and surfacing the single highest-leverage opportunity.
The output is not a score. It is a priority. Your steepest curve right now is X. Here is the one habit that will move you up it. When that habit is automatic, here is what to do next.
Each assessment is anchored to published meta-analyses and dose-response data. Curve parameters and confidence levels are explicit. The methodology is iterated on as the evidence base grows and as user data informs calibration.
A model is a model. Translating population-level effect sizes to individual curves involves judgment. Self-report has well-known limits, and we are upfront about them. Domains interact in ways the current architecture treats as independent — exercise improves sleep, sleep affects nutrition, social connection buffers stress. Some curves are anchored to stronger evidence than others; we mark that explicitly rather than smoothing it over.
We also believe these limits are far less serious than the alternative — which is no structure at all. An imperfect prioritization built on the best available evidence is more useful than the current default: an undifferentiated firehose of advice that nobody can act on.
Samuel Preston showed in 1975 that the relationship between investment and life expectancy follows a predictable curve of diminishing returns. The same principle applies at the individual level — but, until recently, nobody had built the tool to act on it.
The Asymptote Principle is a missing layer in the longevity ecosystem: a scientific framework for answering not “what should I do?” but “what should I do first?” It is, in essence, what happens when you apply structure to the research that already exists.
The asymptote is the theoretical maximum we approach but never reach. The journey toward it is what matters. And the first steps are always the steepest.
1. Preston SH. The changing relation between mortality and level of economic development. Population Studies. 1975;29(2):231-248.
2. Mold JW, Blake GH, Becker LA. The law of diminishing returns in clinical medicine. J Am Board Fam Med. 2010;23(3):371-375.
3. Lee DH et al. Long-term leisure-time physical activity intensity and all-cause and cause-specific mortality. Circulation. 2022;146(7):523-534.
4. Wang F et al. A systematic review and meta-analysis of 90 cohort studies of social isolation, loneliness and mortality. Nature Human Behaviour. 2023;7:1307-1319.
5. Holt-Lunstad J et al. Loneliness and social isolation as risk factors for mortality: a meta-analytic review. Perspectives on Psychological Science. 2015;10(2):227-237.
6. Meta-analysis. Imbalanced sleep increases mortality risk by 14–34%. GeroScience. 2025.
Per-domain evidence libraries — including the full dose-response data, hazard ratios, and curve calibration sources for each pillar — are maintained alongside each assessment's methodology page rather than reproduced here.
This is a living document. Curve parameters and domain weights will be updated as the evidence base grows and as user data informs calibration.